Authors: Shahzad Hassan (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA; 2 Division of Cardiovascular Medicine, Brigham and Women’s Hospital, Boston, MA), Alexander J. Blood (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA; 2 Division of Cardiovascular Medicine, Brigham and Women’s Hospital, Boston, MA; 4 Harvard Medical School, Boston, MA), David Zelle (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA), Sanjay Kumar (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA), Kavishwar Wagholikar (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA), Daniel Gabovitch (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA), Christopher P. Cannon (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA; 2 Division of Cardiovascular Medicine, Brigham and Women’s Hospital, Boston, MA; 4 Harvard Medical School, Boston, MA), Naomi Fisher (3 Division of Endocrinology, Diabetes and Hypertension, Brigham and Women’s Hospital, Boston, MA; 4 Harvard Medical School, Boston, MA), Benjamin M. Scirica (1 Accelerator for Clinical Transformation, Brigham and Women’s Hospital, Boston, MA; 2 Division of Cardiovascular Medicine, Brigham and Women’s Hospital, Boston, MA; 4 Harvard Medical School, Boston, MA)
Categories: Article, Hypertension management, Remote monitoring, Digital healthcare, Long-term outcomes, Population health management
Source: Hypertension (Dallas, Tex. : 1979)
Authors: Shahzad Hassan, Alexander J. Blood, David Zelle, Sanjay Kumar, Kavishwar Wagholikar, Daniel Gabovitch, Christopher P. Cannon, Naomi Fisher, Benjamin M. Scirica
Hypertension (HTN) is a major cardiovascular risk factor, yet traditional care often results in suboptimal blood pressure (BP) control at the population level. We implemented a remote HTN management program that monitored home BP and titrated medications per algorithm. This study assessed the program’s long-term effects by examining participants’ office BP up to 42 months post-enrollment.
Participants of remote HTN program were categorized into four 1) Enrolled-maintenance (achieved goal home BP of ≤130/80 mm Hg), 2) Enrolled-early exit (left before achieving goal BP), 3) Education-only (lifestyle modifications and medications compliance) and 4) White coat HTN group (high office BP but normal home BP). Office BP readings of participants were collected up to 42 months post-enrollment. A linear mixed-effects regression model estimated mean BP levels and studied factors associated with above-goal SBP in the maintenance group.
Office BP readings from 3,601 participants (mean age 61 ± 11 years; 57% female; 60% white; 52% ASCVD) were extracted from EHR and analyzed. All groups sustained office BP below their qualifying values (p <0.001) over 42 months. In the maintenance group, 89.7% of participants maintained SBP at goal, compared to 63.5% in the early exit group, 69.4% in the education-only group and 90.7% in the whitecoat HTN group. Age >50 years was associated with above-goal SBP in the maintenance group.
Participants who achieved BP control through the remote HTN program maintained goal SBP in 90% of cases up to 42 months post-enrollment. These findings highlight the long-term benefits of remote, intensive management programs for effective HTN control.
Hypertension (HTN) is highly prevalent in the general population^1,2^ and is a leading cause of cardiovascular (CV) diseases, contributing significantly to high morbidity, mortality and economic burden^3,4^. Despite the availability of numerous therapeutic options, blood pressure (BP) control among the general population has remained stubbornly suboptimal with traditional care^5^.
Recent technological advancements enable remote monitoring and delivery of HTN care, demonstrating promising results and scalability^6–8^. Telehealth utilization can also increase outreach to rural areas with limited access to healthcare facilities^9^ and those with limited English proficiency, thereby reducing HTN care disparities among different racial and ethnic groups^10^.
While randomized clinical trials have shown the CV benefits from intensive BP control^11,12^, follow-up studies have revealed that BP levels often rise after the active intervention period, highlighting the importance of consistent and sustained management strategies^13^. Although remote HTN management programs can achieve improved BP control during short-term follow-up^14^, the long-term effect of these interventions on BP trajectories after the completion of intervention remains underexplored.
In this study, we conducted a secondary analysis of a remote HTN intervention program, led by pharmacists and patient navigators, which was implemented within the Mass General Brigham (MGB) healthcare network. This program successfully delivered algorithm-based HTN care to patients with elevated baseline BP. The primary objective of this study was to examine the long-term trajectory of office BP in participants following enrollment in the remote HTN intervention. A secondary objective was to identify participant characteristics associated with elevated mean office SBP after achieving goal BP during the remote HTN management phase.
The data that support the findings of this study are available from the corresponding author upon reasonable request.
The structure, methodology and results of remote hypertension management Program have been previously published^6,8,15^. The Accelerator for Care Transformation (ACT) team at Brigham & Women’s Hospital (BWH) and MGB developed a platform to remotely manage poorly controlled hyperlipidemia, HTN and elevated kidney risk in patients with type 2 diabetes mellitus with or at higher risk of atherosclerotic cardiovascular disease (ASCVD), in the patients within the MGB health network^6,16,17^. Potential participants were identified through electronic health record (EHR) and direct provider referrals.
Eligible participants in the remote HTN management program were adults aged ≥ 26 to 80 years with documented uncontrolled HTN, defined as systolic blood pressure (SBP) ≥ 130 mmHg and/or diastolic BP (DBP) ≥ 80 mmHg on at least two of the last three routine office visits within last 18 months, or at least one office SBP ≥ 130 mmHg or DBP ≥ 80 mmHg in the last 18 months with provider referral, or an average 24-hour ambulatory BP ≥ 130/80 mmHg. All participants were followed within the MGB healthcare system. Exclusion criteria included pregnancy or breastfeeding, congestive heart failure with reduced ejection fraction, severe aortic stenosis, bilateral renal artery stenosis, estimated glomerular filtration of <30 ml/min/1.73 m^2^, orthostatic hypotension, and terminal medical condition like malignancy.
All eligible patients were contacted via telephone by navigators. A total of 182,857 were identified as potentially eligible based on EHR data, including BPs and clinical history; 5,046 were contacted, and 3,658 were enrolled. Patients provided verbal consent for participation in the clinical program; written consent was waived as the program was considered a quality improvement project prescribing FDA-approved medication, as per recommendation by professional societies^18,19^. The investigators received approval from the Institutional Review Board (IRB) for database analysis.
Enrolled participants were provided with digitally connected (Bluetooth or cellular enabled) home BP monitors and were educated on their use by program navigators. A program baseline BP was established by recording two morning and two evening home BP readings over six consecutive days. Participants who declined medication titration received education on medication compliance and lifestyle modifications. Non-licensed patient navigators were trained by clinical pharmacists and cardiologists, implemented the care algorithm, while pharmacists prescribed and titrated medications using collaborative drug management agreements, supervised by program physicians.
Participants who achieved the goal mean home BP of ≤130/80 mm Hg or a close approximation (within 1 to 2 mm Hg), based on 2017 ACC/AHA guidelines and calculated from 1 week of home BP readings (two readings each in the morning and evening, minimum of 12 readings), entered the ‘maintenance phase,’ defined as the cessation of active follow-up for BP readings and medication titration by the program. The program collected 424,482 home BP values and issued 15,047 new prescriptions to enrolled participants for a mean of 2.5 medication changes per participant. A mean (SD) of 13.1 (2.9) phone calls were made per participant to reach maintenance. The primary outcome of the study was mean change in home SBP and DBP with remote intervention. After program completion, HTN management was transitioned back to the participant’s primary care team, which included either the primary care physician (PCP) or sub-specialist (cardiologist, nephrologist, or endocrinologist) within the MGB network. The study was conducted from January 2018 to July 2021.
The participants who consented to enroll, transmit BP readings via a wireless BP cuff, and receive medication titration as per the algorithm to achieve the goal BP were “Enrolled” group, of these, participants who achieved goal home BP ≤130/80 were defined as the “Enrolled – Maintenance” group. Within Enrolled cohort, a group of participants “Enrolled – early exit,” did not achieve goal BP due to loss of follow-up, refusal to transmit BP readings, or withdrawal from the program. The “Education only” group did not agree to receive medication titration and transmit home BP readings, hence did not receive home BP device. They were provided with access to audio-visual educational content created for the program, focusing on dietary, lifestyle, and medication adherence. This group served as a non-randomized control group. The group of participants who met the qualifying BP criteria but were at goal on home BP measurements before initiation of the program were termed the “White coat hypertension group”. This group did not transmit further home BP readings or receive pharmacologic intervention.
Baseline sociodemographic characteristics – including age, sex, race and clinical characteristics such as diabetes, ASCVD, hyperlipidemia, congestive heart failure with preserved ejection fraction (CHF-pEF), obstructive sleep apnea (OSA), chronic kidney disease (CKD) were identified from EHR.
Area Derivation Index (ADI) was utilized as an indicator of the socioeconomic status of individual participants. The ADI is a standardized score based on the census variable that reflects an area’s socioeconomic disadvantage, incorporating 17 measures of employment, income, housing quality, and education from the American Community Survey^20–22^. The ADI rankings are indexed to the characteristics of each state within the United States, with ranking from 1 (lowest level of socioeconomic disadvantage) to 10 (highest level of socioeconomic disadvantage).
For this study, routine office BP readings from 6 months before up to 42 months post-enrollment date were extracted from EHR retrospectively. These measurements were obtained during participants’ routine office visits, taken by staff using usual clinical practice techniques and available equipment, without implementation of a standardized research protocol. If multiple BP readings were recorded during a single office visit, an average was calculated and reported for that visit. Readings from emergency room visits, urgent care visits and inpatient admissions were excluded from this analysis. Data was collected between May 2023 to June 2023. This study did not include the home BP readings collected during the intervention phase of the remote HTN program.
In routine clinical practice, office BP measurement techniques vary widely^23,24^ compared to the landmark hypertension trials^25^ where variation could be lessened using standardized protocols^26,27^. This often leads to overestimating BP levels in routine clinical practice, with consequent overtreatment^28^. Routine office BP measurements are susceptible to a “white coat effect”, less accurate and correlating poorly to ambulatory BP measurement^29^. For this study, we defined the goal office SBP ≤140 mm Hg, in alignment with the threshold used in the Healthcare Effectiveness Data and Information Set^30^ performance measure for controlling high BP. This goal is supported by the National Committee for Quality Assurance^30,31^ as an achievable and safe threshold in routine clinical settings, ensuring consistency with widely accepted quality measures for HTN management.
Baseline characteristics of participants were reported as frequencies and proportions for categorical variables and means with standard deviation (SD) or medians with interquartile range (IQR) for continuous variables.
Estimated mean BP values with 95% confidence intervals at 12- and 42-months post-enrollment were calculated using linear mixed effect model with interaction between time since enrollment (in months) and study groups being the fixed effect and participants as the random effects. Changes in estimated mean office BP after enrollment were calculated from each group’s qualifying values and compared between groups at 12- and 42 months post-enrollment. We plotted the trajectory of office SBPs of all study groups post-enrollment using linear mixed effect model, which included the time since randomization and study group interaction as fixed effect and participants as the random effect. The model employed B-splines at 6-months intervals up to 42-months post-enrollment to capture non-linear trend (Figure 1).
A mixed-effects logistic regression model was created to identify variables associated with above-goal mean SBP (>140 mm Hg) in the maintenance group at 42 months post-enrollment. Predictor variables included age group, race, gender, presence of ASCVD, OSA, CKD, diabetes, hyperlipidemia, CHF-pEF, ADI rank and type of insurance plan. To account for intra-patient correlations due to repeated measurements within individuals, we applied clustered standard errors by participant identifier. Statistical significance was set at a p-value of <0.05. All analyses were conducted using R (The R Foundation).
A total of 57,475 office BP readings were extracted from the EHRs of 3,601 program participants, yielding a median (IQR) of 12 (7–20) BP readings per participant. The participants were enrolled for the mean (SD) of 6.0 (5.6) months in the remote HTN management program.
At enrollment, the mean (SD) age was 61.1 (11.7) years, with 71% participants aged 50–75 years, 57% females, 60.2% white and 39.8% with ADI rank 4–6 (middle group). Hyperlipidemia was the most frequently occurring medical comorbidity in all participants (76%) followed by ASCVD (51.9%) and diabetes mellitus (32.9%) (Table 1). Each participant had a median (IQR) of 12 (9–16) BP readings available in EHR post-enrollment over a follow-up period of 42 months.
At 12 months post-enrollment, participants in the maintenance group achieved the largest reduction in mean office SBP, with a change of −9.6 mm Hg (95% CI, −10.4 to −8.8 mm Hg) from their qualifying SBP, compared to −4.0 mm Hg (95% CI, −4.6 to −3.3 mm Hg) in the early exit group and −2.8 mm Hg (95% CI, −4.5 to −1.1 mm Hg) in the education-only group (p <0.001). The white coat hypertension group also experienced a reduction of −9.4 mm Hg (95% CI, −10.5 to −8.3 mm Hg), despite not receiving pharmacologic intervention (Table 2, Figure S1).
At 42 months post-enrollment, the maintenance group sustained the largest cumulative reduction in SBP, with a change of −10.5 mm Hg (95% CI, −11.2 to −9.7 mm Hg). This was followed by a change of −7.5 mm Hg (95% CI, −8.5 to −6.5 mm Hg) in the white coat hypertension group and −5.5 mm Hg (95% CI, −6.1 to −4.9 mm Hg) in the early exit group. The education-only group showed the smallest overall change, with a reduction of −3.0 mm Hg (95% CI, −4.5 to −1.5 mm Hg). Similar trends were observed in diastolic BP (DBP), with the maintenance group consistently achieving the greatest reductions at both 12 and 42 months (Table 2, Figure S2).
At 12-months post-enrollment, mean office SBP was at goal (≤140 mm Hg) in 89.4% (95% CI, 87.3% to 91.2%) of participants in the maintenance group compared to 57.8% (95% CI, 55.4% to 60.3%) participants in the early exit group, 70.0% (95% CI, 64.1% to 75.4%) participants in the education only group and 95.7% (95% CI, 93.7% to 91.1%) participants in white coat hypertension group.
At 42-months post-enrollment, 89.7% (95%CI, 87.6% to 91.5%) participants in the maintenance group, 63.5% (95% CI, 61.2% to 65.8%) in the early exit group, 69.7% (95% CI, 63.7% to 74.6%) participants in the education only group and 90.7% (95% CI, 88.0% to 92.8%) in white coat hypertension group maintained their mean office SBP at goal (Table 3).
In multivariate regression analyses, participants aged 50–75 years (OR 1.52, 95% CI: 1.36–1.70, p < 0.001) and those over 75 years (OR 1.81, 95% CI: 1.58–2.08, p < 0.001) were significantly more likely to have above-goal mean SBP in maintenance groups at 42 months post-enrollment. Conversely, ASCVD was associated with better SBP control, favoring maintenance of at-goal mean SBP (OR 0.86, 95% CI: 0.81–0.92, p = 0.038).
Other factors, including female sex, race, insurance type, ADI rank, and comorbidities such as OSA, CKD, Type 2 diabetes, hyperlipidemia, and CHF with preserved EF, did not show statistically significant associations with above-goal SBP in the maintenance group (Figure 2).
This study provides key insights into long-term BP control after completing a discrete time-limited remote HTN management program. Notably, approximately 90% of participants who achieved goal SBP during the remote HTN management program maintained this control for up to 42 months post-enrollment. Given that over half of the participants of the program also had underlying ASCVD and hyperlipidemia and 32% had underlying diabetes, sustained at goal SBP levels could significantly contribute to reducing the risk of adverse CV outcomes and mortality associated with uncontrolled HTN.
Consistent with our model of care delivery, remote HTN management programs led by pharmacists or trained nurses have successfully managed BP of diverse patient populations^32–34^. In previous studies where participants received usual care versus pharmacist-led remote BP interventions, the latter group achieved better BP control during a 12-month intervention phase and sustained these effects during a 6-month follow-up period^14^. However, data on the longer-term BP outcomes following completion of such interventions have been inconsistent. Our study adds to the overall experience by following the participants and studying their office BP trajectory up to 42-months post-enrollment and showing that participants who successfully completed the remote medications titration achieved a greater sustained decline in the mean BP compared to those who exited the program early or received education only.
Our secondary analysis also examined the association between the characteristics of participants who could not sustain BP control after achieving maintenance in the program. We found the strongest association between poor SBP control and age above 75 years. This finding aligns with the overall rising BP trend in the US population^5^, particularly in individuals aged 60 years and older^35^. Given that the United States population is aging^36^ and associated poor control of BP with age while receiving traditional care, the disease burden of HTN is expected to rise not only in the US but worldwide^37^ leading to a surge in the cardiovascular disease prevalence by 2060^38^. These projections underscore the need for innovative approaches to HTN control, particularly in delivery of care to the elderly population.
Remote delivery of HTN care can help achieve BP control in the older population however this could be challenging due to limited digital literacy, lack of digital access at home^39^, cognitive impairment^40^ and absence of digital care framework for patients in the long term care facilities^41–43^. Our remote hypertension management program utilized cellular-enabled BP cuffs for ease of setup and data transmission to secure servers, mitigating the challenge of manual recording. Additionally, direct phone calls were the primary mode of communication between program navigators and participants, eliminating the need for extensive digital infrastructure and resources.
ASCVD was associated with maintenance of SBP at goal at 42 months in maintenance groups, which could be due to close follow up of these patients with their care providers and improved adherence with the anti-hypertensive regimen.
A higher proportion of participants in the early exit group were black and Hispanic and belonged to areas with higher ADI rank. These factors may have contributed to lower adherence rates due to barriers such as financial constraints or cultural differences. Addressing these disparities within remote HTN management programs could involve implementing targeted strategies, such as culturally tailored educational materials and navigators, and proactive outreach to improve engagement and retention.
Overall, the findings of our study extend the current understanding of the long-term impact of remote HTN interventions, highlighting the potential for improving BP control, scalability, and the importance of individual factors such as age and comorbid conditions in these improvements. These insights can inform the design of future interventions, potentially enhancing their effectiveness and tailoring to the needs of older patients and those with higher prevalence of cardiovascular diseases and risk factors. Furthermore, they underscore the necessity for continuous albeit less intensive management strategies for high-risk patients beyond the duration of intervention to maintain the achieved benefits.
Our data are encouraging regarding the cost of a population health management program, whereby a time-limited remote intervention lasting up to six months could lead to sustained BP control for over 3 years. For healthcare systems aiming to improve BP control at a population level, investing in remote HTN management programs may yield significant long-term benefits, both in terms of health outcomes and cost savings. This suggests that remote care models not only enhance clinical outcomes but can also serve as a sustainable and cost-efficient approach for long-term HTN management.
This study demonstrates that time-limited remote HTN management programs can produce sustainable long-term benefits, with nearly 90% of participants maintaining goal SBP levels for up to 42 months post-intervention. These findings suggest that such programs can serve as an effective and scalable model for population-level BP control, particularly as healthcare systems address the rising burden of hypertension among aging populations. Tailoring remote interventions to meet the specific needs of older adults—such as improving digital literacy and overcoming cognitive or access-related barriers—will be crucial for maximizing impact. The observed association between ASCVD and sustained BP control emphasizes the importance of continuous integration between remote care programs and clinical follow-up to enhance adherence. Improved engagement of primary care physicians, who are at the forefront of treating patients with HTN, and optimization of referral systems can significantly enhance the enrollment process in such programs, leading to improved BP outcomes. Efforts to address the socioeconomic and cultural barriers, particularly among minority groups and those from disadvantaged areas, will be critical for equitable access and improving retention in such programs. As healthcare systems shift toward value-based care, remote HTN programs can offer a promising and cost-effective approach to improve clinical outcomes and address the growing cardiovascular burden in a sustainable manner.
BP readings for this analysis were obtained from the EHR routine office visits documented in the EHR. These readings exhibit high variability within and between participants which could be due to factors such as the time of the day BP is measured^44^, inter-operator variability and white-coat hypertension effect^45,46^. However visit-to-visit variability in BP is common^47^. In this study, each participant had a median of 12 office BP readings post-enrollment, and we reported mean BP among the study population consisting of a large sample size, enhancing the generalizability of our results. Additionally, the qualifying BP had an average of at least three office readings prior to enrollment to account for the variability.
Information about SBP control after the completion of the remote HTN program phase was limited to routine office BP values extracted from the EHR, which are known to have a poor concordance with ambulatory BP measurements. Nevertheless, the office BP values of the participants in this study followed the trend of home BP values during the remote HTN management phase, providing an adequate estimation of their BP trend even after the completion of remote HTN intervention. Our study did not account for changes in the severity of the existing medical conditions or development of new comorbidities, which could require changes in the medication regimen and impact the BP control during the observational phase. We also do not have data on adherence to the anti-hypertensive regimen in the observational phase. However, our study population was large and diverse, receiving standardized care within MGB system, and we observed BP control across all groups. All participants were invited from the MGB health network; however, they are representative of the Massachusetts population in general so the results of this study can be generalizable. Within the enrolled group, 64% of participants did not achieve the goal BP and left the program early but received at least 1 titration, which, along with education, may have contributed towards the improved long-term BP. The white coat hypertension group demonstrated a reduction in BP despite not receiving the pharmacologic intervention or formal education. We hypothesize that the initial positive screening and subsequent introduction to the hypertension program may have prompted participants to adopt lifestyle modifications independently. Qualifying office BP readings in our study cohort were close to 140/80 mm Hg. This reflects the characteristics of the population identified and enrolled in our program while actively receiving care within a large urban healthcare system. While this may limit the generalizability of our findings to patients with more severely elevated office BP, we still demonstrated the potential for remote HTN management programs to sustain BP control even among patients with less pronounced baseline elevation. We emphasize the need for further studies evaluating the effectiveness of such programs in populations with higher baseline BP. We acknowledge that participation in home BP monitoring program may have introduced the selection bias, as individuals who opted into the program belonged to urban areas, were likely more motivated and engaged with their healthcare providers, evident by their qualifying BP. This may have contributed to the high rates of sustained BP control observed in observational phase. Due to lack of control group and observation design of this study, causality of sustained BP control cannot be established, and the findings should be interpreted cautiously. Despite these limitations, our study represents a critical step toward understanding the longer-term impact of digital care interventions in managing HTN at a population level. Further research is needed to determine whether remote HTN interventions can reduce cardiovascular and all-cause mortality over longer-term follow-up.
This study underscores the effectiveness of remote HTN management programs in achieving and sustaining goal BP levels. Notably, among participants who successfully completed the program, 90% maintained their mean SBP at goal up to 42-months post-enrollment. Remote HTN interventions can play an important role in long-term BP control, especially in older patients and those with existing or higher cardiovascular risks. Our findings advocate for the broader implementation of remote HTN management programs due to their scalability. These programs can be pivotal in addressing the growing burden of HTN, particularly when the traditional care models have not effectively improved BP control at the population level. Future studies should validate these results in broader populations, asses impact on cardiovascular outcomes, and optimize remote programs to enhance accessibility, sustainability, and equity in HTN management.